COLUMBIA GLACIEB, 268 making the hill a nunatak. At least one great south-flowing tributary enters the main glacier above this distributary, and there may be two other north or northwest tributaries. The west side has two pronounced embayments, the northernmost of which may contain a tributary glacier while the southern seems from one point of view to be simply a lateral valley into the mouth of which the Columbia Glacier margin has bulged westward. The broad lower portion of Columbia Glacier is one of the largest and most beautiful in Alaska, terminating in a superb ice cliff from which icebergs are continually being discharged. The order of magnitude of this ice tongue may be seen by examining PL CI in which the whole city of Washington, D. C/, has been drawn carefully to scale upon the surface of Columbia Glacier. One who has walked much in Washington can appreciate the magnitude of a surface of snowy-white, severely-crevassed ice twice as wide as from the Capitol to the White House and with a terminal ice cascade nearly as long as from the Navy Yard to the Naval Observatory, and varying in height from 100 to 250 feet. There are medial moraines near each side of the lower glacier and a broad lateral moraine along the eastern margin, which is lobate and determined in position by several rock hills 500 to 1200 feet high. The distributary east of the nunatak is fronted by a low rock ridge and has its terminus covered with ablation moraine. The west lateral moraine of this distributary swings out into the ice tongue and becomes medial in the upper main glacier. The glacier is moving rapidly and is severely crevassed from side to side, though less broken within the area of lateral moraines. For purposes of description Columbia Glacier may be divided into (a) the main, glacier surface, (b) the lobate eastern margin, (c) the eastern ice cliff, (d) the Heather Island terminus, (e) the main ice cliff, and (f) the western margin. These will be described in inverse order. The Western Margin. Gilbert observed in 1899 l that "at the western margin of the main ice cliff, where the glacier crowded against a steep slope, there was a belt of bare rock, from 200 to 300 feet broad, between the ice and the forest." A photograph taken by him from a rocky point about half a mile south of the western edge of the glacier front shows the conditions in 1899 and ten years later the National Geographic Society's expedition was able to reoccupy this exact site (G. Fig. 81), as Grant had done earlier in the same year. Our 1909 photograph is reproduced with that of Gilbert in PL CII and shows the precise change that had taken place in the ten year interval. There was considerable lateral spreading and thickening, for, as the glacier advanced, it had broadened so that the ice margin had pushed forward two or three hundred feet and once more reached the forest, into which it had not yet advanced very far on August 23, 1909. This is shown by the presence, in both pictures, of the same distinctive lone tree and the adjacent trees at the top of the hill between the ice edge and the forest^ The glacier had thickened one hundred and fifty feet or thereabouts. The lateral spreading may have been more than 200 feet, for the glacier surely continued to recede between 1899 and the time when the 1909 advance began and it must, therefore, have readvanced this distance in addition to the 200 feet. On July 5,1910, the junior author found that this lateral spread had continued, covering the site of a marginal stream near the glacier terminus and crowding still farther westward into the forest. The amount of spread in the ten month interval was about i Gilbert; G. K, Harrixnan Alaska Expedition, Vol. ffi, Glaciers and Glaciation, 1004, pp. 76-77.